There's a chip shop on the front at Eyemouth I've watched for twenty years. Every season the owner adds something — halloumi fries, a breakfast menu, Mondays now open too. The queue has done the opposite: it used to run past the arcade on a wet Tuesday, now it barely reaches the door. Ask him how business is and he'll say turnover's up. He measures effort — menu items, opening hours — and has quietly stopped measuring what the effort is supposed to produce.

All of which brings me to the British Horseracing Authority's fixture list, which has run the Eyemouth chip shop's business plan for a decade, on a sport I happen to like.

Fewer Horses, More Races

Proform Racing, who've logged every British run since 1997, sent over the numbers this week. A flat horse still runs about as often as it always has — 5.2 to 5.6 times a season, unmoved for thirty years, the Covid year aside. Jumpers: 3.4 to 3.75, equally static. What's moved is how long each horse sticks around. A flat horse debuting in 1999–2003 ran 15.8 times before retiring; one debuting in 2004–2008 ran 14.6; one debuting 2014–2018 ran 13.2 — a career 10 per cent shorter than twenty years ago, 16 per cent shorter than at the turn of the century. The share reaching 30-plus starts has fallen from 15 per cent to 11. Jump careers have barely moved since 2004. (Some of the flat drop is exports, not retirements — Proform can't see a horse that's left the country.)

The pool of available runners isn't shrinking because horses run less. It's shrinking because each one lasts less time. Into that shrinking pool, the BHA has poured a decade of extra fixtures and all-weather cards. You don't need a maths degree for what happens next. You need a chip shop.

The Idea Too Sensible to Print

A correspondent of mine, Andrew, wrote to a national paper suggesting that if the BHA is serious about turnover, it should put someone from the betting industry in charge of Race Planning. Not consulted — in charge. The letter didn't run. To most letters editors, handing a racing job to a bookmaker apparently reads as a joke rather than a policy — which is the whole problem in miniature: the people who understand what drives turnover are a punchline to the people setting the programme meant to produce it.

The Cliff Nobody's Allowed to Measure

Here's what Proform can show, because they hold the prices. As fields shrink, favourites shorten — hard. From 2023–2025, an 18-plus runner field returns an average Betfair favourite of 5.8, winning 17 per cent of races. At 12–14 runners: 4.0, 28 per cent. At 8–9: 3.2, 35 per cent. At 6–7: 2.8, 38 per cent. At five or fewer: 2.3, winning nearly half the time, odds-on in a third of those. The market's calibrated at every band — the price genuinely tracks the horse's chance. But the contraction steepens below eight runners, exactly where a British race also drops from three each-way places to two. Two mechanisms, same postcode. Forty-two and a half per cent of British races now sit under that line.

Does turnover fall steadily as fields shrink, or off a cliff below eight? Andrew suspected the second. Proform's honest answer: they can't say, because only bookmakers and the Levy Board hold turnover at race level, and the Levy Board publishes totals, not fixtures. I asked the Levy Board for exactly that data myself. I was told, informally, that bookmaker confidentiality prevented it — an interesting stance for a board funded by taking a cut of punters' money, since it means those punters can never be told what happened to it. The mechanism is proven. The consequence isn't — because the people who could prove it have decided not to let anyone try. Racing, as a sport, has never had to be financially transparent. On this evidence, it shows.

Who Benefits From Not Knowing

Racecourses are largely paid through media-rights deals that don't move much whether a race draws eighteen runners or five — so a lost field doesn't cost the racecourse much; the cost lands downstream, on the levy pool and on bookmakers stuck trading short-priced markets nobody wants. The BHA sets the fixture list inside a governance structure where racecourses hold real sway over the schedule that pays them. Nobody needs to be acting in bad faith. It only takes nobody with the numbers being in the room — and nobody in the room asking why they aren't.

Credit where it's rare in this sport: Proform, a commercial subscription business with every reason to keep its methods to itself, lets any subscriber build their own rating in pounds, on the handicapper's own scale, and test it against every British result since 2018 — something the regulator has never bothered to make possible in public. A subscription service has made racing's numbers more legible than the body statutorily responsible for the sport. That should embarrass someone.

The Detonator Down the Corridor

There's a second reckoning coming, and it isn't a horse race. The Reuben brothers will eventually sell ARC — Arena Racing Company — which, between it and Jockey Club Racecourses, controls most of the tracks this fixture list runs on. JCR haven't the money to buy a rival. Ascot can't, structurally. The bookmakers, who once wanted a stake in racecourse ownership, now see the whole sport as a rounding error beside their casino apps. Which leaves private equity — who don't need to love the sport, only to be better at long division than the people currently running it. On this evidence, that's not a high bar.

A Problem That Only Gets Worse

None of this eases off from here. The foal crop is shrinking as fast as careers are, which means the pool of runners feeding this fixture list will keep thinning for years. Field size, and the sheer amount of racing being staged, are about to become far more sensitive questions than they are today — and there is no visible strategy for a shrinking population of racehorses meeting a fixture list built for a larger one.

What Would Actually Fix It

The ask, addressed to people who won't read it: the BHA should publish a fixture-level turnover model by field-size band before adding another race to a schedule it can't currently defend. The Levy Board should release the granularity it already collects, not the granularity that's easiest to defend in a board meeting. And somebody should put a person who's actually priced a horse race in charge of the fixture list.

None of that needs new legislation or a single extra penny — just people willing to check whether their own schedule works. Andrew wrote to the Post and got nowhere. I doubt this does much better. But the chip shop in Eyemouth is still open on Mondays, and the queue still isn't there. At some point, somebody has to stop counting the griddles and start counting the customers.

Figures on runs, careers and Betfair starting prices: Proform Racing / RaceMetrics, British racing 1997–2025 (career data), 2023–2025 (price/field-size data, clear favourites) and 2026 to 7 September (share of races under eight runners), unless otherwise stated.

RaceMetrics is a data partner of Cutting Comment.